• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.414-423
• /
• 2020

In this paper, we propose a Dilated Convolution Gate Linear Unit (DCGLU) to mitigate the lack of sparsity and small receptive field problems caused by the segmentation map extraction process in sound event detection with weak labels. In the advent of deep learning framework, segmentation map extraction approaches have shown improved performance in noisy environments. However, these methods are forced to maintain the size of the feature map to extract the segmentation map as the model would be constructed without a pooling operation. As a result, the performance of these methods is deteriorated with a lack of sparsity and a small receptive field. To mitigate these problems, we utilize GLU to control the flow of information and Dilated Convolutional Neural Networks (DCNNs) to increase the receptive field without additional learning parameters. For the performance evaluation, we employ a URBAN-SED and self-organized bird sound dataset. The relevant experiments show that our proposed DCGLU model outperforms over other baselines. In particular, our method is shown to exhibit robustness against nature sound noises with three Signal to Noise Ratio (SNR) levels (20 dB, 10 dB and 0 dB).

• Journal of the Korean association of regional geographers
• /
• v.5 no.2
• /
• pp.119-132
• /
• 1999

The Korean Peninsula, jutting southward from the Asian mainland into the Pacific Ocean, is surrounded by numerous bays and islands on three sides. The study area of this research is the coastal area surrounding the Sacheon Bay, which is located in the mid-southern tip of the peninsula. Historically, the bay region took the role of the main gate leading to the western part of Kyung-nam Do(Province) due to the fact that the bay is the nearest coast to Chinju, the central city of the province. The Sacheon Bay had provided important sea routes from the old days until the early twentieth century. Because of the bay's sea routes, the bay region has been militarily of great significance. Recently, road/rail/highway/airplane transportation functions have become gradually increased but sea transportation function has been decreased in the region. As the Sacheon Bay was rich in fishery resources, dozens of fishing villages in the bay region depended on fishery and marine product industries for long years. But the inflow of the fresh water used to cause serious damage to the fishing industry in the bay region after a drainage canal from the Nam River was constructed in 1969. As a result, the industry has dwindled gradually for last decades. Recently, a couple of industrial parks were constructed along the coast and many manufacturing factories were built on the parks. It has been expected that many jobs lost in fishery and agriculture are replaced by new opportunities in industry. More than half of the workers employed in the parks' factories commute from the city of Chinju. It is asserted in this paper that the transportation, fishery, manufacturing, and tourism functions of the Sacheon Bay region are closely associated with. The characteristics of the region can be well understood through the functional changes and associations.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.10A
• /
• pp.1223-1229
• /
• 2004

A variable-gain low-voltage low noise amplifier MMIC operating at 5GHz frequency band is designed and implemented using the ETRI 0.5$\mu\textrm{m}$ GaAs MESFET library process. This low noise amplifier is designed to have the variable gain for adaptive antenna array combined in HIPERLAN/2. The feedback circuit of a resistor and channel resistance controlled by the gate voltage of enhancement MESFET is proposed for the variable-gain low noise amplifier consisted of cascaded two stages. The fabricated variable gain amplifier exhibits 5.5GHz center frequency, 14.7dB small signal gain, 10.6dB input return loss, 10.7dB output return loss, 14.4dB variable gain, and 2.98dB noise figure at V$\_$DD/=1.5V, V$\_$GGl/=0.4V, and V$\_$GG2/=0.5V. This low noise amplifier also shows-19.7dBm input PldB, -10dBm IIP3, 52.6dB SFDR, and 9.5mW power consumption.

• Journal of rehabilitation welfare engineering & assistive technology
• /
• v.10 no.3
• /
• pp.237-242
• /
• 2016

The major advantage of slip-ring technology in Spiral CT is that it facilitates continuous rotation of the x-ray tube, so that volume data can be acquired from a patient quickly. Not only for such a fast scan, but also for the dose reduction purpose, high signal-to-noise ratio and fast data acquisition system is required. In this study, we have built a multi-channel photodetector and multi-channel data acquisition system for CT application. The detector module consisted of CdWO4 crystal and Si photodiode in 16 channels. For the performance test of the preamplifier stage, both the transimpedance and switched integrator types are optimized for the photodetector modules. Switched integrator showed better noise performance in the limited bandwidth which is suitable for the current CT application. The control sequence for data acquisition and 20 bit ADC is designed with VHDL(Very High Speed Integrated Circuit Hardware Description Language) and implemented on FPGA(Field Programmable Gate Array) chip. Our Si photodiode detector module coupled to CdWO4 crystal showed comparable signal with other commercially available photodiode for CT. Switched integrator type showed higher SNR but narrower bandwidth compared to transimpedance preamplifier. Digital hardware is designed by FPGA, so that the control signal could be redesigned without hardware alteration.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.15 no.1
• /
• pp.39-50
• /
• 2003

This study aims to figure out water quality impacts due to sudden disturbances of sediments during dike construction in land reclamation for the northern part of the Siwhaho Lake where heavily deteriorated settlements from upstream loadings are outstanding. We exploit a 3-D water quality model CE-QUAL-ICM combined with a hydrodynamic model TIDE3D. Simulations are done accounting water-sediment interaction in a 4-$\sigma$ layers. Long-term simulation for 1-year shows that bottom layers around the disturbance location are only affected and marks very high concentration. Complete vertical mixing appears at least 5km apart to downward due to complex effects of geometry, bathymetry and river inflows. It should be addressed that existing condition of the Siwhaho Lake stands for high concentration of COD and TP in winter and spring due to relatively high incoming loadings, however the effect of sediment disturbances yields reverse phenomena, i.e., impacts of dike construction arise greatly in summer and fall. Refined grid system consisting of 150m${\times}$150m rectangular grid, which is doubled system compared to previous study (Suh et al.,2002), gives affordable results by reducing flux differences through a cell especially in front of gate.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.8 no.2
• /
• pp.430-439
• /
• 2004

This paper describes a design of 8192/2048-point FFT/IFFT processor (CFFT8k2k), which performs multi-carrier modulation/demodulation in OFDM-based DVB-T receiver. Since a large size FFT requires a large buffer memory, two design techniques are considered to achieve memory-efficient implementation of 8192-point FFT/IFFT. A hybrid structure, which is composed of radix-4 single-path delay feedback (R4SDF) and radix-4 single-path delay commutator (R4SDC), reduces its memory by 20% compared to R4SDC structure. In addition, a memory reduction of about 24% is achieved by a novel two-step convergent block floating-point scaling. As a result, it requires only 57% of memory used in conventional design, reducing chip area and power consumption. The CFFT8k2k core is designed in Verilog-HDL, and has about 102,000 Bates, RAM of 292k bits, and ROM of 39k bits. Using gate-level netlist with SDF which is synthesized using a $0.25-{\um}m$ CMOS library, timing simulation show that it can safely operate with 50-MHz clock at 2.5-V supply, resulting that a 8192-point FFT/IFFT can be computed every 164-${\mu}\textrm{s}$. The functionality of the core is fully verified by FPGA implementation, and the average SQNR of 60-㏈ is achieved.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.12 no.3
• /
• pp.312-320
• /
• 2019

In this paper, we design a 512Kb eFlash IP using 110nm eFlash cells. We proposed eFlash core circuit such as row driver circuit (CG/SL driver circuit), write BL driver circuit (write BL switch circuit and PBL switch select circuit), read BL switch circuit, and read BL S/A circuit which satisfy eFlash cell program, erase and read operation. In addition, instead of using a cross-coupled NMOS transistor as a conventional unit charge pump circuit, we propose a circuit boosting the gate of the 12V NMOS precharging transistor whose body is GND, so that the precharging node of the VPP unit charge pump is normally precharged to the voltage of VIN and thus the pumping current is increased in the VPP (boosted voltage) voltage generator circuit supplying the VPP voltage of 9.5V in the program mode and that of 11.5V in the erase mode. A 12V native NMOS pumping capacitor with a bigger pumping current and a smaller layout area than a PMOS pumping capacitor was used as the pumping capacitor. On the other hand, the layout area of the 512Kb eFlash memory IP designed based on the 110nm eFlash process is $933.22{\mu}m{\times}925{\mu}m(=0.8632mm^2)$.

• 한국노년학
• /
• v.28 no.4
• /
• pp.953-968
• /
• 2008

This research was performed to recognize anxiety level of older people, to find out anxiety level after or before regular exercises, and to find out how much anxiety level could be lowered by the sports activities. This research selected a population among the students who enrolled at sport activity classes (e, g., physical exercises, gate ball, volume dance) at L, S, J, Y (a), Y (b), D older people college located in Seoul and Gyoenggi area. 200 of the population were chosen as candidates by convenience sampling, nonprobability sampling. Considering the age of the candidates, total 200 cases were collected by using two methods (survey and interview) simultaneously. The data assessment was made by SPSS 12.0 Version. Error tolerance in statistics is .05. The data was analyzed by using frequency analysis, paired t-test and independent t-test. After analyzing anxiety awareness level after and before regular exercises, anxiety of older people are classified by four factors. First, anxiety for loss is 3.756 (M=3.756) before exercises, while 1.942 (M=1.942) after exercises. Second, fear for aging is 3.443 before exercises and 2.243 after exercises. Third, anxiety for physical appearance is 3.253 before exercises and 2.310 after exercises. Finally, anxiety caused by psychological insecurity is 3.060 before exercises, while 1.666 after exercises. Error of tolerance for all factors falls within .001. Anxiety score after exercises is lower than that of before exercises for every factor as well. As a result, regular physical exercises appeared to reduce anxiety level of older people.

• Smart Media Journal
• /
• v.8 no.2
• /
• pp.58-65
• /
• 2019

Automatic text summarization is a process of shortening a text document by either extraction or abstraction. The abstraction approach inspired by deep learning methods scaling to a large amount of document is applied in recent work. Abstractive text summarization involves utilizing pre-generated word embedding information. Low-frequent but salient words such as terminologies are seldom included to dictionaries, that are so called, out-of-vocabulary(OOV) problems. OOV deteriorates the performance of Encoder-Decoder model in neural network. In order to address OOV words in abstractive text summarization, we propose a copy mechanism to facilitate copying new words in the target document and generating summary sentences. Different from the previous studies, the proposed approach combines accurate pointing information and selective copy mechanism based on bidirectional RNN and bidirectional LSTM. In addition, neural network gate model to estimate the generation probability and the loss function to optimize the entire abstraction model has been applied. The dataset has been constructed from the collection of abstractions and titles of journal articles. Experimental results demonstrate that both ROUGE-1 (based on word recall) and ROUGE-L (employed longest common subsequence) of the proposed Encoding-Decoding model have been improved to 47.01 and 29.55, respectively.

• The Journal of Engineering Geology
• /
• v.23 no.2
• /
• pp.105-115
• /
• 2013

Increasing the river cross-section by barrage construction causes rises in the average river water levels and discharge rates in the rainy season. The time series patterns for groundwater levels measured at 23 riverside monitoring wells along the lower Nakdong River are compared for two cases: before and after water-filling at the Changnyeong-Haman Barrage. Monthly average groundwater levels indicate a distinct increase in groundwater levels in the upstream riverside close to the barrage. River-water level management by barrage gate control in August, during the rainy season, resulted in a 0.1 m decrease in groundwater levels, while water-filling at the barrage in December caused a 1.3 m increase in groundwater levels. The results of hierarchical cluster analysis indicate that seven groundwater monitoring wells and river water levels were in the same group before barrage construction, but that this number increased to 14 after barrage construction. Principal component analysis revealed that the explanation power of two principal components corresponding to river fluctuation, PC1 and PC2, was approximately 82% before barrage construction but decreased to 45% after construction. This finding indicates that the effect of the river level component that contributes to change in groundwater level, decreases after barrage construction; consequently, other factors, including groundwater pumping, become more important. Continuous surveying and monitoring is essential for understanding change in the hydrological environment. Water policy that takes groundwater-surface water interaction into consideration should be established for riverside areas.